Date: | July 16, 2007 |
---|---|
Location: | DC 1304 |
Time: | 1:30 |
Chair: | Curtis Luk |
Date: | July 23, 2007 | July 30, 2007 | August 13, 2007 | August 20, 2007 |
---|---|---|---|---|
Location: | DC 1304 1:30 | DC 2303(Lab) 1:30 | DC 1304 1:30 | DC 1304 1:30 | Chair: | Jaime Ruiz |
Christine Szentgyorgyi |
Martin Talbot |
Mike Terry |
Technical Presentation: | Yingbin Liu |
Curtis Luk |
Jaime Ruiz |
Christine Szentgyorgyi |
Yi Lin |
Title : Nonuniform Segment Compression of Motion Capture Data
Abstract: I am going to introduce applying a scale-space nonuniform segmentation by anisotropic diffusion to the application of motion capture data compression. The basic idea is each degree of freedom of a motion clip is smoothed by an anisotropic diffusion process and then divided into segments at the zero crossings of the second derivative in the smoothed data. Then such a segment is approximated by a cubic Bezier curve. The approximation produces a hierarchical lists of cubic control points, which can be further compressed by a hierarchical coding method. The coding method uses the points in the coarse scales to predict the ones in the finer scales. In this end, I will compare our work with wavelet compression methods which have the best compression rates to date. |
---|
Also see other Math and CS postings.